HighwayGraph: Modelling Long-distance Node Relations for Improving General Graph Neural Network
This addresses a key limitation in GNNs for graph-structured data applications, though it is incremental as it builds on existing shallow GNN architectures.
The paper tackles the problem of modeling long-distance node relations in Graph Neural Networks (GNNs), which conventional GNNs perform poorly on due to limited-layer propagation, by proposing HighwayGraph, a model-agnostic training framework that uses shallow GNNs with implicit and explicit relation modeling, achieving consistent and significant improvements over four representative GNNs on three benchmark datasets.
Graph Neural Networks (GNNs) are efficient approaches to process graph-structured data. Modelling long-distance node relations is essential for GNN training and applications. However, conventional GNNs suffer from bad performance in modelling long-distance node relations due to limited-layer information propagation. Existing studies focus on building deep GNN architectures, which face the over-smoothing issue and cannot model node relations in particularly long distance. To address this issue, we propose to model long-distance node relations by simply relying on shallow GNN architectures with two solutions: (1) Implicitly modelling by learning to predict node pair relations (2) Explicitly modelling by adding edges between nodes that potentially have the same label. To combine our two solutions, we propose a model-agnostic training framework named HighwayGraph, which overcomes the challenge of insufficient labeled nodes by sampling node pairs from the training set and adopting the self-training method. Extensive experimental results show that our HighwayGraph achieves consistent and significant improvements over four representative GNNs on three benchmark datasets.